Multiple Target Tracking Based On Vo-Vo Filter

With the continuous development of sensor technology,multi-target tracking technology has been widely used in many fields,increasingly highlighting its important role.Target tracking refers to the actual status,trajectory,and number of targets obtained from the measurement information.In order to quickly form the trajectory and effectively track the target,this paper mainly studies the multi-target tracking algorithm based on Vo-Vo filter.This algorithm provides a new research direction for target formation track management.In this paper,we first review the principles of Vo-Vo filter algorithm.It includes the concept of label random finite set,GLMB filter,Delta-GLMB filter,LMB filter and so on,and describes the internal relations,differences and filtering process of these algorithms in detail.The relevant algorithms of Vo-Vo filter are deduced from the theory of delta-GLMB filter,and there is no detailed literature to introduce its implementation.Aiming at this problem,we use Gauss mixture to realize the delta-GLMB filtering algorithm,and verify its tracking performance by simulation.The simulation results show that the ?-GLMB filter can effectively track the target,accurately estimate the number of targets,and form the track.But the disadvantage is that the calculation is large and the filtering process takes a long time.Aiming at the shortcomings of ?-GLMB filter,this paper adopts a new implementation method: joint prediction and updating to process the filtering process.This joint strategy requires only one truncation process during each iteration,eliminating the inefficiencies of truncated processing during prediction and update,respectively.After using the truncated index weighted sum to rank the ?-GLMB filter,the computational complexity is also the third power.Therefore,this article truncates the treatment using Gibbs sampling to reduce dimensionality.Finally,the algorithm is implemented by Gaussian mixture,and its tracking performance is verified by simulation.Simulation results show that the optimized tracking algorithm can effectively track the target and greatly reduce the amount of computation.